Ultrasonic Rate Measurements in Two-Phase Bubble Flow

One of the most important tools in well testing and production logging is the downhole flowmeter. Unfortunately, existing spinner and tracer tools have limited accuracy in low-rate single phase flow and multiphase flow situations. Two experimental studies investigated the use of ultrasonic methods to measure flow rates under these conditions. The experiments showed that the ultrasonic transmission method can be used for accurate measurements of low-rate single phase water flow. The ultrasonic reflection method provided reasonably accurate bubble velocities in two-phase bubble flow regime under laminar flow conditions. Experiments using ultrasonic transmission provided higher accuracy and better sensitivity in measurement of low-rate single phase water flow than conventional methods. The one-inch square flowcell device investigated in this study used the transmission properties of ultrasonic waves to measure the volumetric rate in single phase water flow. The flow cell was installed in a vertical flow loop. Water also passed through a section where the weight of water was recorded accurately against time, thus giving another rate measurement for cross-checking purposes. The device-measured flow rates were compared with actual rates obtained by the weight measurement. The device was accurate down to a flow rate of 1.0 gallon/min, which is the showest flow rate achievable in this loop. Such a low flow rate would normally be difficult to measure using a conventional spinner tool. Experiments using ultrasonic reflection examined how accurately an experimental flow tool measured bubble velocity in a two-phase bubble flow regime. Measurements were made in a graduated, transparent pipe with an inside diameter of 5.50 inches. The tool diameter is 3.25 inches. Gas and oil were bubbled through a stagnant column of water, and bubble velocities were measured. To cross-check the accuracy of tool measurements, bubble motion was also recorded using a video camera. The video frame had a clock, with a resolution of one hundredth of a second, providing time information. Tool-measured bubble velocities were compared with actual velocities determined by a frame-by-frame playback of video data. The tool readings were linearly related to the actual velocities in the range from 0.38 ft/sec to 1.30 ft/sec, although the velocities registered by the tool were consistently in error. This velocity information, in combination with the void fraction data, will provide the dispersed phase volumetric flow rate. The result is significant because, in contrast to the currently employed spinner and tracer tools, this technique can provide us with flow rates for individual phases in a two-phase bubble flow situation.